Of the hundreds of moons in our solar system, only one has its own magnetic field. That moon is Ganymede — Jupiter’s largest satellite, bigger than the planet Mercury — and scientists have puzzled over it since NASA’s Galileo spacecraft first confirmed the field existed in the 1990s. A new study in Science Advances, published in May 2026, proposes an explanation that may finally make sense of a decades-long contradiction. It is a compelling idea, though not yet the final word.
The study, “Powering Ganymede’s Dynamo with Protracted Core Formation,” is led by Kevin Trinh, a planetary scientist at Caltech. The core idea: Ganymede’s metallic centre may not be fully formed yet. It may still be slowly pulling itself together — and the act of forming may be exactly what powers the magnetic field.
A contradiction four billion years in the making
Magnetic fields like Ganymede’s are generated by churning liquid metal deep inside a planet or moon. Several worlds have them: Earth, Mercury, and all four giant planets — Jupiter, Saturn, Uranus and Neptune. Among the solar system’s hundreds of moons, however, Ganymede stands alone. As Trinh puts it: “Callisto is similar in size and density, but it has no obvious evidence of a dynamo. Why are they so different?”
For decades, two bodies of research have tried to answer that question — and arrived at opposite conclusions. Studies of how Ganymede formed suggest it came together too cold and too slowly to have built a metallic core early on. But models of how its magnetic field works have assumed exactly that: a core that formed billions years ago, much like Earth’s, cooling down ever since. Both pictures cannot be right at the same time.
Trinh’s team decided to take the formation side seriously. What if Ganymede really did start cold, without a finished core, and has been warming up and building one ever since? This “cold-start” scenario is the foundation of the new model.
The warming-driven mechanism
Using computer simulations, the researchers modelled Ganymede’s interior evolving slowly from a cold, unformed state. Heated gradually over billions of years by radioactive decay and Jupiter’s gravitational pull, liquid metal would have begun sinking toward the centre — and, according to the model, may still be doing so today.
That slow inward migration, the team argues, is enough to stir the forming core and generate a magnetic field. They call it a “warming-driven dynamo,” a different idea from the standard picture, where a long-finished core loses heat and that temperature difference drives the field. Here, it is the act of building the core, not cooling it, that powers everything.
This is physically plausible partly because of what Ganymede’s core is probably made of. An iron-sulfide mix — Fe-FeS, in shorthand — melts at lower temperatures than pure iron. In what geologists call a “sub-eutectic” regime, where the interior sits below the temperature at which the mixture would melt entirely, the alloy can still partially liquefy as it warms. The denser iron-rich liquid then sinks, separating from the lighter sulfide-rich material. That gradual separation is the process the paper refers to as differentiation, and it is what the model identifies as the source of the churn. A cold-formed moon with Fe-FeS chemistry would not need to have been especially hot inside; it only needs to keep warming.
What the model does and does not say
This study does not prove the warming-driven mechanism is how Ganymede actually works. What it shows is that the idea is viable — consistent with both the cold-formation history and the magnetic field we see today. As Trinh has noted, the results “do not rule out a cooling-driven dynamo at Ganymede.” They add a new option that has not been on the table before.
The study is also a computer model, not a direct look inside the moon. Models depend on assumptions about composition, temperature history, and internal structure, and several of those assumptions are still uncertain for Ganymede. This is the opening of a new line of inquiry, not the end of one.
What JUICE may tell us
The clearest test may come from the European Space Agency’s JUICE mission — the JUpiter Icy Moons Explorer — due to arrive at Jupiter in 2031. JUICE will measure Ganymede’s magnetic field in far more detail than Galileo could, and will also map the moon’s gravity and internal structure. That data may eventually be enough to distinguish between the two explanations, or it may raise questions that neither model has anticipated.
If the warming-driven picture holds up, Ganymede would be unusual in a specific way: a moon that has spent 4.5 billion years still in the process of becoming itself.
